Kelsen and Hart on the Normativity of

Law

Torben Spaak∗

1 Introduction …………………………………………………………………... 398

2 The Problem About the Normativity of Law ……………………………….. 398

3 Kelsen’s Account of the Normativity of Law ……………………………….. 402

3.1 Kelsen’s Theory of Law ………………………………………………… 402

3.2 The Basic Norm as Hypothesis …………………………………………. 404
3.3 The Basic Norm as Fiction ……………………………………………... 405
3.4 The Basic Norm and the Normativity of Law ………………………….. 406

4 Hart’s Account of Law’s Normativity ………………………………………. 407

4.1 Hart’s Theory of Law …………………………………………………… 407

4.2 The Rule of Recognition ………………………………………………... 408
4.3 The Rule of Recognition and the Normativity of Law …………………. 410

5 The Trump Thesis and the Significance of Strictly Legal Normativity …... 412

∗ I would like to thank Uta Bindreiter for helpful comments on this article, and Åke Frändberg,
Lars Lindahl, and Lennart Åqvist for discussing with me some of the problems touched upon
in this article. I would also like to thank Robert Carroll for checking my English. As always,
the author alone is responsible for any remaining mistakes and imperfections.

© Stockholm Institute for Scandianvian Law 1957-2010

398 Torben Spaak: Kelsen and Hart on the Normativity of Law

1 Introduction

The problem about the normativity of law – that is, the problem of accounting
for the nature of the legal ought, the law’s normative force, or, if you will, the
nature of legal reasons for action – is in my view the most serious if not the only
serious question facing legal positivists. The problem, I have argued elsewhere,
is that although legal positivists can account for law’s normativity in (what I
shall call) the strictly legal sense, but not in (what I shall call) the moral sense,
such an account is difficult to combine with the thesis that law necessarily
claims to trump moral and other reasons for action.1 But, one may wonder, if
normativity in the strictly legal sense is so problematic, why have first-rate legal
thinkers like Hans Kelsen and Herbert Hart expended so much energy in trying
to account for the normativity of law in this sense? Were they perhaps really
concerned with normativity in the moral sense? I don’t think so. I think they
were concerned with normativity in the strictly legal sense. And in this article I
am going to consider Kelsen’s and Hart’s analyses of the normativity problem in
order to show that both Kelsen and Hart were indeed concerned with normativity
in the strictly legal sense. I am also going to argue that their accounts can be
combined with a qualified version of the thesis that the law necessarily claims to
trump moral and other reasons for action, while suggesting that the normativity
of law in the strictly legal sense is not so important a characteristic of law as
Kelsen and Hart seem to have thought.
I begin by presenting the problem about the normativity of law (Section 2). I
then proceed to consider Kelsen’s and Hart’s analyses of this problem (Sections
3-4). Having done that, I consider the possibility of combining the strictly legal
conception of law’s normativity with a qualified version of the trump thesis, and
add a few words about the significance of strictly legal normativity (Section 5).

2 The Problem About the Normativity of Law2

As is well known, natural lawyers and legal positivists hold opposing views
about the nature of law. While they tend to agree that law is a system of norms,
they disagree about the relation between law thus conceived and morality. The
debate, as I see it, concerns concept formation at the most fundamental level in
the study of law: How should we understand and shape the concept of law? And
what role, if any, should moral considerations play in such concept formation?
Natural law theory understood as a theory of law takes positive law, that is,
law laid down by humans for humans, to be inherently and genuinely normative,
necessarily conferring genuine rights and imposing genuine obligations. And it
accounts for this binding force by asserting that positive law is conceptually
connected with moral values like justice and the common good. Generally
speaking, the idea is that there is a higher law, which we can discover by using

1 See Spaak, Torben, Legal Positivism, Law’s Normativity, and the Normative Force of Legal
Justification, in Ratio Juris Vol. 16:4 2003, p. 469-85.
2 The text in this section can be found, more or less verbatim, in Spaak, Legal Positivism,
supra note 1, p. 471-2, 478-81.

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 399

our reason and which confers binding force on positive law, if and insofar as the
latter is in keeping with the former. More specifically, natural law theory asserts
(i) that there is a conceptual connection between law and morality, and (ii) that
moral values and standards exist independently of people’s beliefs and attitudes.3
On this analysis, the moral authority of law is part of the concept of law, and the
thesis that an unjust law cannot be legally valid, i.e., cannot be a law at all (lex
injusta non est lex), turns out to be a corollary to (i).
Legal positivism is a general and descriptive theory of law of the type
advanced by scholars like John Austin,4 Hans Kelsen,5 Alf Ross,6 H. L. A. Hart,7
Joseph Raz,8 and Neil MacCormick & Ota Weinberger,9 not a theory telling the
judge how he should decide hard cases or when civil disobedience is justified.10
Underlying, though neither entailing nor entailed by, legal positivism is meta-
ethical noncognitivism, according to which moral claims have no cognitive
meaning.11 Legal positivism thus conceived could perhaps be described as a
meta-theory, a theory about theories of law, because it aims to lay down
requirements that any adequate theory of law must meet.12 Since legal positivists
usually exclude from the study of law questions concerning the moral value of
law, they tend to describe law in terms of formal features, saying for example
that it is a “specific social technique of a coercive order.”13
Now the problem about the normativity of law, as I have said, concerns the
nature of the legal ought or law’s normative force, or, if you will, the nature of
legal reasons for action. Philosophers tend to conceive of normativity in general
as that which is common to the normative (right, wrong, duty) and the evaluative
(good, bad) in regard to theoretical as well as practical questions.14 We are not
concerned with normativity in general, however, but with legal normativity; and
I take legal normativity to be stronger than other types of normativity – ex-

3 See, e.g., Aquinas, Thomas, Summa Theologiae (R. J. Henle ed.) 1993, Q 90, Art. 2, C. &
Art. 4, C.; A. P. d’Entrèves, Natural Law, London 1951, p. 85; Moore, Michael S., Law as a
Functional Kind, in Natural Law Theory (Robert P. George ed.) Oxford 1992, p. 189-92;
Radbruch, Gustav, Rechtsphilosophie , 7th. ed., Stuttgart 1950, p. 353.
4 Austin, John, The Province of Jurisprudence Determined, Indianapolis/Cambridge 1954.
5 Kelsen, Hans, General Theory of Law and State, Cambridge, Massachusetts 1945; Kelsen,
Hans, Reine Rechtslehre, 2d. ed., Wien 1992 [1960].
6 Ross, Alf, On Law and Justice, Berkeley & Los Angeles 1959.
7 Hart, H. L. A., The Concept of Law, Oxford 1961; Hart, H. L. A., Essays on Bentham,
Oxford 1982.
8 Raz, Joseph, The Authority of Law, Oxford 1979, Ch. 3; Raz, Joseph, Authority, Law and
Morality, The Monist Vol. 68 1985, p. 295-324.
9 MacCormick, Neil & Weinberger, Ota, An Institutional Theory of Law, Dordrecht 1986.
10 See also Gardner, John, Legal Positivism: 5 1/2 Myths, in American Journal of Jurisprudence
Vol. 46 2001, p. 199 and Hartney, Michael, Dyzenhaus on Positivism and Judicial
Obligation, in Ratio Juris Vol. 7 1994, p. 48-51.
11 See MacCormick, Neil, Legal Reasoning and Legal Theory, 2d. ed., Oxford 1994, p. 5;
Kelsen GTLS, supra note 5, p. 13-14.
12 See Raz, AL, supra note 8, p. 39.
13 Kelsen GTLS, supra note 5, p. 19.
14 See Dancy, Jonathan, Editor’s Introduction, Normativity (Jonathan Dancy ed.) Oxford 2001,
p. 1.

© Stockholm Institute for Scandianvian Law 1957-2010

400 Torben Spaak: Kelsen and Hart on the Normativity of Law

cepting moral normativity of course. The reason is that law necessarily claims to
trump moral and other reasons for action.15 That is to say, law does not, except
in extreme cases, recognize as legally relevant conflicts between legal and moral
reasons for action. From the viewpoint of the courts, acts of civil disobedience
and conscientious objection cannot be accepted, unless there is a legal norm
authorizing the judge to take certain moral arguments into account.16
Now, as many writers have noted, the obvious way to account for the
normativity of law is to argue that having a legal right or obligation is having a
special kind of moral right or obligation. Ronald Dworkin,17 Lon Fuller,18 and
Aleksander Peczenik,19 among others, have analyzed the normativity of law
along these lines. This conception, which I shall refer to as the moral conception
of law’s normativity, is attractive, because it makes it clear why we should care
about our legal rights and obligations and why we should obey the law. For on
this analysis, a person who is legally obligated to do X is necessarily morally
obligated to do X; and that explains why we should be interested in our legal
rights and obligations, and it also explains (roughly) what it means to have an
obligation to obey the law.
But not everyone believes that having a legal right or obligation is having a
special kind of moral right or obligation. Some maintain instead that having a
legal right or obligation is having a sort of strictly legal right or obligation, that
is, a legal right or obligation sui generis.20 H. L. A. Hart’s critique of John
Austin’s theory of law illustrates (what I shall refer to as) the strictly legal
conception of law’s normativity. Hart rejects Austin’s sanction theory of legal
obligation because he believes it obliterates the important distinction between
being obligated to do something and being obliged (or forced) to do it.21 To
bring out the inadequacy of Austin’s analysis, he considers a situation in which a
person is ordered by a gunman to hand over his money. As Hart sees it, the
victim may be obliged – but not obligated – to hand over the money. This
distinction is important to Hart because, he says, “[l]aw surely is not the gunman
situation writ large, and legal order is surely not to be thus simply identified with
compulsion.”22

15 When I say that the law makes a claim, I mean that judges – the law’s spokesmen – make that
claim when they act on behalf of the law. See Alexy, Robert, Law and Correctness, Legal
Theory at the End of the Millennium (M. D. A. Freeman ed.) 1998, p. 205; Soper, Philip, The
Ethics of Deference, Cambridge 2002, p. 7.
16 This is nicely illustrated by the Supreme Court’s reasoning in the Swedish case NJA 1982 s
621.
17 Dworkin, Ronald, Taking Rights Seriously, 2d. ed. London 1978, Ch. 4.
18 Fuller, Lon L., Positivism and the Fidelity to Law: A Reply to Professor Hart, Harvard Law
Review Vol. 71 1958, p. 644-57.
19 Peczenik, Aleksander, Vad är rätt?, Stockholm 1995, p. 527-8.
20 The term ‘strictly legal’ is my own invention. See Spaak, Torben, (Review of) Rex Martin, A
System of Rights Theoria, Vol. 61 1995, p. 80. It has not been used by Kelsen, Hart or other
leading legal positivists.
21 Hart, CL, supra note 7, p. 79-88.
22 Hart, H. L. A., Positivism and the Separation of Law and Morals, Harvard Law Review Vol.
71 1958, p. 603.

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 401

Hart maintains that what is missing in Austin’s theory is the idea of a rule.23
According to Hart, we need the idea of a rule in our analysis of the concept of a
legal obligation. For to say that someone has an obligation (legal or moral) to
perform an action is to assume a back-ground of rules that makes certain
behavior standard, and to apply a rule to that person and his behavior.24 Such
duty-imposing rules, Hart tells us, are “conceived as binding independently of
the consent of the individual bound.”25
Now according to Hart, a statement that a person has a legal obligation refers
to an action that is “due from or owed by” the person having the obligation, in
the sense that it “may be properly demanded or extracted from him according to
legal rules or principles regulating such demands for action.”26 On this analysis,
when a judge states that someone has a legal obligation to pay his taxes, say, he
may mean to “speak in a technically confined way,” that is, he may mean to
speak from within a legal institution that he is committed as a judge to maintain;
and in so doing he draws attention to what may be legally demanded of the
person having the obligation.27 Although the judge may morally approve of this
obligation, his moral approval is not part of the meaning of his legal statement.28
The strictly legal conception of law’s normativity is problematic, however.
The main problem is that it seems to be impossible to combine it with the thesis
that law necessarily claims to trump moral and other reasons for action.29 For
how can a judge, who conceives of legal reasons for action as something that
necessarily trumps moral and other reasons for action, also think of himself as
speaking from within an institution in the sense indicated by Hart? Hart himself
notes that ”to many it will seem paradoxical, or even a sign of confusion, that . .
. I should argue that judicial statements of the subject’s legal duties need have
nothing directly to do with the subject’s reasons for action.”30 I fear that I am
one of the many.31
Let us now turn to consider Kelsen’s and Hart’s analyses of law’s normativity
in order to see how they understood the problems involved and how they
proposed to solve them. Let us begin with Kelsen’s analysis.

23 Hart, CL, supra note 7, p. 78.

24 Id. p. 83.
25 Id. p. 168.
26 Hart, EB, supra note 7, p. 159-60.
27 Id. p. 266.
28 Id. p. 266.
29 For a version of this thesis, See Joseph Raz, Practical Reason and Norms, 2d. ed., Princeton
1990, p.151-2
30 Hart, EB, supra note 7, p. 267.
31 See also Raz, Joseph, Hart on Moral Rights and Legal Duties, in Oxford Journal of Legal
Studies Vol. 4 1983, p. 129-31.

© Stockholm Institute for Scandianvian Law 1957-2010

402 Torben Spaak: Kelsen and Hart on the Normativity of Law

3 Kelsen’s Account of the Normativity of Law

3.1 Kelsen’s Theory of Law

The Pure Theory of Law is a general theory of law that conforms to the
requirements of legal positivism.32 As such, it aims to understand the law as it is,
not as it ought to be, and its method is structural analysis.33 More specifically, it
provides us with a set of fundamental legal concepts – such as ‘legal system,’
‘norm,’ ‘right,’ ‘duty,’ ‘sanction,’ and ‘imputation’ – that we can make use of
when trying to understand and describe the law in a scientific manner.34 We
might say that the Pure Theory aims to lay down the theoretical basis for other
legal disciplines, such as contract law, constitutional law, legal history,
comparative law, etc.35
The Pure Theory conceives of law as a system of norms,36 which norms
function as schemes of interpretation in light of which we can view human
behavior and other natural events.37 The structure of such a system is described
by Kelsen as a Stufenbau, that is, a structure of norms on different levels where
norms on a higher level authorize the creation of norms on a lower level.38 On
Kelsen’s analysis, a norm is the meaning of an act of will directed at the
behavior of another. As such it expresses – just like the orders issued by a Mafia
boss – a subjective ought. Legal norms differ, however, from the orders issued
by the Mafia boss in that they also express an objective ought: that the act in
question ought to be performed not only from the viewpoint of the person
positing the norm, but also from the viewpoint of the person whose behavior the
norm regulates, and from the viewpoint of a neutral third party.39
To say that a legal norm is valid, Kelsen explains, is to say that it exists, and
to say that it exists is to say that it ought to be obeyed or applied, that it has
binding force.40 To say that a valid legal norm expresses an objective ought is
just another way of expressing the same idea. Kelsen maintains, in keeping with
the separation thesis, that legal validity is conceptually independent of morality:
“[t]here is no kind of human behavior that, because of its nature, could not be

32 Kelsen, GTLS, supra note 5, p. xiii; Kelsen, Hans, On the Pure Theory of Law, Israel Law
Review Vol. 1 1966, p. 5.
33 Kelsen, RR II, supra note 5, p. 112.
34 Kelsen, Hans, The Function of the Pure Theory of Law, Law: A Century of Progress 1835 to
1935. Vol. 2 1937, p. 231; Kelsen, GTLS, supra note 5, p. xiii-xiv.
35 Kelsen, Function , supra note 34, p. 232.
36 Kelsen RR II, supra note 5, p. 215-21. See also Kelsen, GTLS, supra note 5, p. 3.
37 Kelsen RR II, supra note 5, p 3-4. See also Kelsen, GTLS, supra note 5, p. 41.
38 Kelsen RR II, supra note 5, p. 228.
39 Id. p. 4-5, 7-8, 45-6; Kelsen, Hans, What is the Pure Theory of Law?, 34 Tulane Law
Review Vol. 34 1959/60, p. 270.
40 Kelsen RR II, supra note 5, p. 9-10. See also Kelsen, GTLS, supra note 5, p. 30, 39. Kelsen
thus treats ‘norm’ and ‘valid norm’ as synonyms, which means that a non-valid norm is not a
norm at all. For the sake of simplicity, I will sometimes allow myself to depart from
Kelsen’s usage and speak of non-valid norms.

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 403

made into a legal duty corresponding to a legal right.”41 He also maintains, in

keeping with the is/ought distinction, that the validity of a given legal norm can
only be explained by reference to the validity of another and higher legal norm.
Thus a norm, n1, is legally valid if, and only if, it was created in accordance with
another and higher legally valid norm, n2, which in turn is legally valid if, and
only if, it was created in accordance with another and higher legally valid norm,
n3, etc.42
We should note here that Kelsen accepts as fundamental and self-evidently
correct the distinction between what is and what ought to be,43 between the
world of is and the world of ought, as he used to say in his earlier writings.44 He
conceives of ‘is’ (Sein) and ‘ought’ (Sollen) as two fundamental and distinct
categories or modes of thought,45 and he takes the meaning of ‘ought’ to be
intuitively clear, expressing “the specific sense in which human behaviour is
determined by a norm.”46 ‘Ought,’ he says, is a simple notion, and it can
therefore not be defined:47 “ebensowenig, wie man beschreiben kann, was das
Sein oder das Denken ist, ebensowenig gibt es eine Definition des Sollens.”48
Law, then, is a normative phenomenon, and as such it must be carefully
distinguished from factual phenomena,49 but also from other normative
phenomena.50 Since this is so, legal scholars can invoke neither (i) empirical
considerations from psychology, sociology, economics, political science, etc.,
nor (ii) normative considerations from ethics, theology, etc. in their analyses of
the law.51 As Kelsen says, the basic methodological aim of the Pure Theory is to
free the study of law from all foreign elements, to avoid methodological
syncretism.52 This is what the purity of the Pure Theory amounts to.
As one might expect, Kelsen rejects John Austin’s command theory of law.53
He maintains instead that a command can be binding only if the commander has
the legal power to issue that command, and that the commander’s legal power
depends on the existence of a legal system that confers on him the requisite legal
power. Hence a gangster’s command that you hand over your money to him
cannot be binding, as there is no valid legal norm conferring legal power on the
gangster to issue such commands. Kelsen concludes that Austin wrongly thought

41 Kelsen, GTLS, supra note 5, p. 113. See also Kelsen, RR II, supra note 5, p. 200-1.
42 Kelsen, RR II, supra note 5, p. 196-7. See also Kelsen, GTLS, supra note 5, p. 110-1.
43 Kelsen RR II, supra note 5, p. 5-6. See also Kelsen, GTLS, supra note 5, p. 36-7, 110-1.
44 Kelsen, Hans, Hauptprobleme der Staatsrechtslehre, 1911, p. 7-10.
45 Id. p. 7-8.
46 Kelsen, GTLS, supra note 5, p. 37. Negatively, Kelsen says that “p ought to do A” means
neither that the speaker or someone else wants p to do A, nor that p will in fact do A. Id. p.
37.
47 Kelsen, RR II, supra note 5, p. 5, note *. He holds that G. E. Moore’s characterization of
‘good’ – that it is a simple notion like ‘yellow’ – applies to ‘ought,’ too. Id. p. 5, note *.
48 Kelsen, Hauptprobleme, supra note 44, p. 7.
49 Id. p. 1-2.
50 Kelsen RR II, supra note 5, p. 61.
51 Id. at 1. See also Kelsen, On the Pure Theory, supra note 32, p. 2-3.
52 Kelsen RR II, supra note 5, p. 1.
53 Kelsen, GTLS, supra note 5, p. 30-2. See also Kelsen, RR II, supra note 5, p. 45-6.

© Stockholm Institute for Scandianvian Law 1957-2010

404 Torben Spaak: Kelsen and Hart on the Normativity of Law

he could derive the binding force of a command from the command itself, when
instead he should have focused on the conditions under which the command is
issued.

3.2 The Basic Norm as Hypothesis

When tracing the validity of a given legal norm through the chain of validity,
one finally arrives at the historically first constitution. Since that constitution
cannot have been created in accordance with another and higher valid norm,
Kelsen terminates the chain of validity by simply presupposing that we ought to
behave in accordance with the historically first constitution.54 He calls this
presupposition the basic norm (die Grundnorm), and explains that it is ”the final
postulate, upon which the validity of all the norms of our legal system
depends.”55 So the basic norm is the tool we use to distinguish between law and
coercion, between being obligated and being obliged,56 which means that it
grounds the normativity of law.
Kelsen emphasizes that although the basic norm refers directly to a specific
constitution in an efficacious legal system,57 the moral value of the legal system
does not enter into consideration when one presupposes the basic norm. On the
contrary, “[i]n der Voraussetzung der Grundnorm wird kein dem positiven Recht
transzendenter Wert bejaht.”58 Accordingly, he maintains that one may well
presuppose the basic norm in a situation where the Mafia has effective control
over a certain geographical area, thus excluding competing coercive orders, and
so consider their coercive order a legal system.59 He admits, to be sure, that we
may ask why we ought to obey the historically first constitution, but points out
that it is characteristic of legal positivism to dispense with religious and moral
justifications of law.60
Although Kelsen seems to take the ontological basis of norms to be the norm-
giver’s will, he characterizes the basic norm as an epistemological device for
conceiving of the legal materials as valid legal norms:
So wie Kant fragt: wie ist eine von aller Metaphysik freie Deutung der
unseren Sinnen gegebenen Tatsachen in den von der Naturwissenschaft
formulierten Naturgesetzen möglich, so fragt die Reine Rechtslehre: wie ist eine
nicht auf meta-rechtliche Autoritäten wie Gott oder Natur zurückgreifende
Deutung des subjektiven Sinns gewisser Tatbestände als ein System in
Rechtssätzen beschreibbarer objektiv gültiger Rechtsnormen möglich? Die

54 Kelsen RR II, supra note 5, p. 197, 203. See also Kelsen, GTLS, supra note 5, p. 115.
55 Kelsen, GTLS, supra note 5, p 115.
56 Kelsen RR II, supra note 5, p. 204-9. See also Kelsen, GTLS, supra note 5, p. 115-7.
57 Kelsen RR II, supra note 5, p. 204.
58 Id. p. 204. See also id. p. 223-4.
59 Kelsen, GTLS, supra note 5, p. 45-9. See also Kelsen RR II, supra note 5, p. 223-4.
60 Kelsen, GTLS, supra note 5, p. 116.

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 405

erkenntnistheoretische Antwort der Reinen Rechtslehre lautet: unter der

Bedingung, daß man die Grundnorm voraussetzt.61
This characterization makes it clear that the act of presupposing the basic
norm is really an act of cognition, not an act of volition, and that therefore the
basic norm is the meaning of an act of thinking, not the meaning of an act of
will.62 It is also in keeping with Kelsen’s view that anyone interested in
conceiving of the law as a system of valid norms – judges, lawyers, legal
scholars, ordinary citizens – may but does not have to presuppose the basic
norm:
Die Grundnorm kann, muß aber nicht vorausgesetzt werden. Was die Ethik
und Rechtswissenschaft von ihr aussagt ist: Nur wenn sie vorausgesetzt wird,
kann der subjektive Sinn der auf das Verhalten anderer gerichteten Willensakte
auch als ihr objektiver Sinn, können diese Sinngehalte als verbindliche Moral-
oder Rechtsnormen gedeutet werden. Da diese Deutung durch die
Voraussetzung der Grundnorm bedingt ist, muß zugegeben werden, daß Soll-
Sätze nur in diesem bedingten Sinne als objektiv gültige Moral- oder
Rechtsnormen gedeutet werden können.63

3.3 The Basic Norm as Fiction

After years of referring to the basic norm as a hypothesis,64 Kelsen changed his
mind in the beginning of the 1960’s, suggesting instead that we think of it as a
fiction as that concept is understood in Hans Vaihinger’s Philosophy of As-If.65
Having maintained for a long time that the basic norm is really the meaning of
an act of thinking, Kelsen now emphasizes that there is an important correlation
between will (Wollen) and ought (Sollen), so that there can be no norm without a
corresponding act of will.66 Accordingly, he explains that presupposing the basic
norm involves presupposing an imaginary authority, over and above the
“fathers” of the historically first constitution, whose act of will has the basic
norm as its meaning.67 But, he points out, this means that the notion of the basic
norm contains a contradiction within itself, as it involves presupposing the
existence of an authority that could not possibly exist.68
Kelsen concludes that the basic norm is best described as a genuine fiction in
the Vaihingerian sense. Following Vaihinger, he conceives of a fiction as an aid
to thought (ein Denkbehelf) to be used when one cannot reach one’s aim of

61 Kelsen RR II, supra note 5, p. 205.

62 Id. p. 205-6.
63 Kelsen, Hans, Allgemeine Theorie der Normen (Kurt Ringhofer & Robert Walter, eds.) Wien
1979, p. 206. See also Kelsen RR II, supra note 5, p. 223-4.
64 See, e.g., Kelsen, GTLS, supra note 5, p. 116.
65 Kelsen, Hans, Die Funktion der Verfassung, (Neues) Forum Vol. 132 1964, p. 585; Kelsen,
On the Pure Theory, supra note 32, p. 6-7; Kelsen, ATN, supra note 63, p. 206-7. See also
Vaihinger, Hans, Die Philosophie des Als Ob (4th. ed.) 1920.
66 Kelsen, Verfassung, supra note 65, p. 585.
67 Id. p. 585.
68 Id. p. 585.

© Stockholm Institute for Scandianvian Law 1957-2010

406 Torben Spaak: Kelsen and Hart on the Normativity of Law

thought (Denkzweck) with the materials available.69 Kelsen’s aim of thought, as

we have seen, is to ground the normativity of the legal system, and he admits
that he can achieve this goal only by introducing a fiction, viz. the basic norm.
Kelsen scholars have been debating whether the change of status of the notion
of the basic norm is an important event in the development of Kelsen’s theory of
the basic norm and, therefore, of the Pure Theory. Iain Stewart, for example,
contends that this change signifies the Pure Theory’s, indeed legal positivism’s,
swan song,70 whereas Richard Tur and Robert Walter are of the opinion that the
change is of little consequence.71 Stewart maintains, more specifically, that if the
concept of a basic norm is a fiction, then everything that follows from it, such as
the concept of a legal system, must be fictional too.72 I agree with Stewart that if
the concept of a basic norm, properly understood, is self-contradictory, then it
has to go. I am not, however, convinced that there is such a close connection
between will (Wollen) and ought (Sollen) as Kelsen assumes. In fact, I am not
even convinced that we need to conceive of the basic norm as a norm. We might
simply think of it as a presupposition that the historically first constitution is
legally valid – and as such it is not contradictory.

3.4 The Basic Norm and the Normativity of Law

We see, then, that Kelsen’s proposed solution to the problem about the
normativity of law, considered within the framework of legal positivism, is to
presuppose the basic norm conceived of either as an hypothesis or as a fiction.
But is this really a solution to our problem? As we shall see, the answer to this
question depends on which conception of legal normativity we have in mind.
Joseph Raz has made an attempt to reconstruct the theory of the basic norm in
its role as an explanation of the normativity of law.73 Raz introduces the notion
of a legal man – a man who accepts all and only the laws of his country as
morally valid, as his personal morality, as it were, because he morally endorses
the basic norm –74 and suggests that legal scholars should adopt this point of
view in “a special professional and uncommitted sense” of ‘adopt.’75 Spe-
cifically, he suggests legal scholars should view the law from the legal man’s

69 Id. p. 585.
70 Stewart, Iain, The Basic Norm as Fiction, Juridical Review 1980, p. 207-8. See also.
Paulson, Stanley L., Kelsen’s Legal Theory: The Final Round, Oxford Journal of Legal
Studies Vol. 12, p. 268-70; Alf Ross, Directives and Norms, 1968, p. 29-33, 156-8; Stewart,
Iain, Kelsen and the Exegetical Tradition, in Essays on Kelsen (Richard Tur & William
Twining, eds.) 1986, p. 131-5.
71 Tur, Richard, The Kelsenian Enterprise, in Essays on Kelsen, supra note 70, p. 167-75;
Walter, Robert, Zum Versuch einer Kritik der Reinen Rechtslehre, Rechtstheorie Vol. 21
1990, p. 148-9.
72 Stewart, The Basic Norm as Fiction, supra note 70, p. 208.
73 Raz, AL, supra note 8, p. 122-45; Raz, Joseph, The Purity of the Pure Theory, Revue
Internationale de Philosophie Vol. 138 1981, p. 441.
74 Raz, AL, supra note 8, p. 140. See also Raz, Purity, supra note 73 p. 451-2.
75 Raz, AL, supra note 8, p. 142-3. Note that Raz uses the verb ‘adopt’ rather than
‘presuppose.’

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 407

point of view “as if it is valid or on the hypothesis that it is . . . but without

actually endorsing it.”76 On this analysis, there is no specifically legal
normativity, but only a “specifically legal way in which normativity can be
considered”,77 in that the validity of the constitution is simply (non-committally)
assumed.78
I find Raz’s analysis illuminating.79 For one thing, it is in keeping with the
notions that legal validity is conditional upon presupposing the basic norm and
that one may, but does not have to, presuppose the basic norm. Moreover, it is in
keeping with the notion that one may presuppose the basic norm even in a
situation of Mafia-rule. But, as should be clear, the theory of the basic norm thus
conceived can only account for the normativity of law in the strictly legal
sense.80 On this reading, Kelsen appears as a positivistic hard-liner much like
John Austin, and we may therefore conclude that Kelsen was indeed concerned
to account for the normativity of law in the strictly legal sense.

4 Hart’s Account of Law’s Normativity

Like Kelsen, Hart conceives of law as a system of norms, the foundation of

which is a single, fundamental norm. And just as Kelsen’s account of law’s
normativity rests ultimately on a fundamental, presupposed norm (the basic
norm), Hart’s account of law’s normativity rests ultimately on a fundamental,
accepted norm, which he calls the rule of recognition. The problem Hart faces is
therefore the same problem as Kelsen faced, viz. to explain how this
fundamental norm can ground the normativity of law.

4.1 Hart’s Theory of Law

Like Kelsen, Hart offers a general theory of law that conforms to the
requirements of legal positivism. Like Kelsen, Hart conceives of law as a system
of norms, or as he says, rules. More specifically, he conceives of law as a system
of primary duty-imposing rules and secondary rules of change, adjudication and
recognition.81 Duty-imposing rules are of course the paradigm of rules, as they

76 Id. p. 157.
77 Id. p 145.
78 Raz, Purity, supra note 73, p. 459.
79 But note that Raz’s analysis has been criticized by a number of legal scholars. See, e.g.,
Bindreiter, Uta U., Why Grundnorm?, Dordrecht 2002, p. 94-5; Paulson, Stanley L., Kelsen
Without Kant, in Öffentliche oder private Moral? Festschrift für Ernesto Garzón Valdès
(Werner Krawietz & Georg Henrik von Wright, eds.) Berlin 1992, p.160-2; Vernengo,
Roberto J., Kelsen’s Rechtssätze as Detached Statements, Essays on Kelsen, supra note 70, p.
99; Wilson, Alida, Joseph Raz on Kelsen’s Basic Norm, American Journal of Comparative
Law Vol. 27 1982, p. 46.
80 James Harris Seems to share this view, though he does not accept Raz’s analysis and does not
speak of ‘strictly legal normativity.’ See Harris, James W., Kelsen’s Pallid Normativity, Ratio
Juris Vol. 9:1 1996, p. 94.
81 Hart, CL, supra note 7, at 91.

© Stockholm Institute for Scandianvian Law 1957-2010

408 Torben Spaak: Kelsen and Hart on the Normativity of Law

directly guide human behavior by giving reasons for action: no normative

system can do without them. The secondary rules, on the other hand, are about
the primary rules in the sense that they are used to identify, and to create,
change, and extinguish primary rules, and to set up legal institutions that apply
the primary rules. Specifically, rules of change confer legal power on persons,
thus enabling them to change legal positions;82 rules of adjudication constitute
courts and other law-applying organs and regulate their activities;83 and the rule
of recognition lays down criteria for the identification of the rules of the
system.84 According to Hart, the introduction of these three types of secondary
rules into a set of primary rules may be considered “the step from the pre-legal
to a legal world.”85

4.2 The Rule of Recognition

The rule of recognition fulfills two important functions. First, it identifies and
ranks the sources of law: legislation, precedent, custom, etc.86 Second, it
constitutes the ultimate source of law’s normativity by imposing a legal duty on
the officials to apply all and only norms that meet the criteria of validity laid
down in it.87 The second function clearly presupposes that normativity can
somehow be transmitted from the highest level of the Stufenbau down to the
lower levels.
To be sure, Hart does not explicitly state that the rule of recognition
constitutes the ultimate source of law’s normativity, that the normative force of a
given legal rule depends on the normative force of the rule of recognition. But he
clearly thinks it does.88 Why else would he insist—as laid down by his theory of
social rules—that the rule of recognition is a rule, as distinguished from a mere
habit, and that a legal rule is valid if, and only if, it has been created in
accordance with another and higher legal rule, and ultimately in accordance with
the rule of recognition?89
The rule of recognition is a customary or, as Hart says, a social rule. In other
words, it is a rule by virtue of being accepted by a certain group of people, viz.
the legal officials. So whereas other legal rules exist in the sense that they meet

82 Id. p. 93-4.
83 Id. p. 94-5.
84 Id. p. 92-3.
85 Id. p. 91.
86 Id. p. 97-107.
87 On this issue, See Dworkin, TRS, supra note 17, p. 48-51; MacCormick, Neil, H. L. A. Hart,
Stanford 1981, p. 110; Raz, Joseph, The Concept of a Legal System, 2. ed., Oxford 1980, p.
199.
88 See Dworkin, TRS, supra note 17, p. 21; Postema, Gerald J., Coordination and Convention
at the Foundations of Law, Journal of Legal Studies Vol. 11 1982, p. 170.
89 Hart has criticized Kelsen for speaking of the validity of the basic norm, asserting that the
rule of recognition “can neither be valid nor invalid but is simply accepted as appropriate for
use . . . .” Hart, CL, supra note 7, p. 105-6. But since Hart takes ‘valid’ to mean ‘satisfies all
the criteria provided by the rule of recognition,’ Id. p. 100, his criticism does not show that he
is indifferent to the normativity of the rule of recognition.

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 409

the criteria of validity laid down in the rule of recognition, the rule of
recognition itself “exists only as a complex, but normally concordant, practice of
the courts, officials, and private persons in identifying the law by reference to
certain criteria.”90 This means that “[f]or the most part the rule of recognition is
not stated, but its existence is shown in the way in which particular rules are
identified, either by courts or other officials . . . .”91
Social rules, Hart explains, have an internal aspect, in addition to the external
aspect that they share with habits, and which consists in a certain regularity of
behavior.92 Accordingly, his account of the normativity of social rules centers on
this internal aspect, or more specifically, on the characteristic pro-attitude
toward the rules among those concerned that he refers to as the internal point of
view. He describes the internal point of view as “a critical reflective attitude to
certain patterns of behaviour as a common standard,”93 which displays itself in
criticism of deviant behavior and in recognition that such criticism is justified,
etc.94
So, on Hart’s theory, there is a Danish (say) rule of recognition if Danish
legal officials (i) display a regular pattern of behavior with regard to the
identification and ranking of the sources of law (the external aspect), and (ii)
have a steady commitment in favor of acting in accordance with this pattern of
behavior (the internal aspect).95 And this rule of recognition constitutes the
ultimate source of normativity of Danish law in that it imposes on the legal
officials a legal duty to apply all and only those rules identified in accordance
with it.
Note that on Hart’s theory, part of the reason for each official to comply with
the rule of recognition is that other officials comply with it.96 As Hart has made
clear, legal officials must view the rule of recognition as a “common standard of
correct judicial decision, and not as something which each judge obeys merely
for his part only.”97 The rule of recognition can therefore be described as a
conventional rule.98
Hart’s rule of recognition differs from Kelsen’s basic norm in several
respects.99 The most important one is that whereas the rule of recognition is a
social rule, the basic norm is merely a presupposition, an idea in the minds of
legal scholars and others. Whereas Hart makes it clear that he grounds law’s

90 Hart, CL, supra note 7, p. 107. Strictly speaking, the rule of recognition is addressed only to
the legal officials. See Raz, CLS, supra note 87, p. 198.
91 Hart, CL, supra note 7, p. 98.
92 Id. p. 54-6.
93 Id. p. 56.
94 Id. p. 56.
95 For an analysis of the place of the rule of recognition in American law, See Greenawalt,
Kent, The Rule of Recognition and the Constitution, Michigan Law Review Vol. 85 1987, p.
621.
96 Hart, H. L. A., The Concept of Law, 2d. ed. (With a postscript edited by Penelope A. Bulloch
& Joseph Raz) Oxford 1994, p. 255.
97 Hart, CL, supra note 7, p. 112.
98 Hart, CL II, supra note 96, p. 255.
99 Hart himself has commented on these differences. Hart, CL, supra note 7, p. 245-6.

© Stockholm Institute for Scandianvian Law 1957-2010

410 Torben Spaak: Kelsen and Hart on the Normativity of Law

normativity in social facts, Kelsen avoids this strategy, as he believes it violates

the is/ought distinction. Kelsen could thus accept the rule of recognition as a
criterion of validity, but not as the source of law’s normativity.

4.3 The Rule of Recognition and the Normativity of Law

We see, then, that Hart’s proposed solution to the problem about the normativity
of law, considered within the framework of legal positivism, is to point to the
rule of recognition. But is this an adequate solution to our problem? As we shall
see, the answer to this question – like the answer to the corresponding question
about the basic norm – depends on which conception of law’s normativity we
have in mind.
Let us begin by asking whether the rule of recognition is best understood as a
moral or as a non-moral rule. Lon Fuller maintains that the rule of recognition
must be a moral rule, because it will have to derive its efficacy “from a general
acceptance, which in turn rests ultimately on a perception that [it is] right and
necessary.”100 The idea, then, is that unless the officials morally approve of the
rule of recognition, they will not look upon it as a standard to be complied with,
or criticize those who do not comply, etc. On this analysis, the internal point of
view is a moral point of view and the rule of recognition is therefore a moral
rule.
To be sure, when accepting the rule of recognition, the legal officials are,
strictly speaking, accepting a customary rule laying down essentially factual
criteria of validity. One might, therefore, be tempted to argue that the rule of
recognition has nothing to do with morality. I doubt, however, whether the
officials can really disregard the actual content and function of the legal system
in question when contemplating whether to accept the rule of recognition. If, for
example, the legal system were grossly immoral, wouldn’t this influence the
legal officials’ acceptance? I think it would. If I am right, accepting the rule of
recognition involves accepting the legal system, and acceptance of the legal
system would seem to be a paradigm case of moral acceptance.
Nevertheless, Hart denies that the legal officials who accept the rule of
recognition need consider themselves morally bound to do so:
Not only may vast numbers be coerced by laws which they do not regard as
morally binding, but it is not even true that those who do accept the system
voluntarily, must conceive of themselves as morally bound to do so, though the
system will be most stable when they do so. In fact, their allegiance to the
system may be based on many different considerations: calculations of long-
term interest; disinterested interest in others; an unreflecting inherited or
traditional attitude; or the mere wish to do as others do. There is indeed no
reason why those who accept the authority of the system should not examine
their conscience and decide that, morally, they ought not to accept it, yet for a
variety of reasons continue to do so.101

100 Fuller, Positivism, supra note 18, p. 639. See also Goldsworthy, Jeffrey D., The Self-
Destruction of Legal Positivism, Oxford Journal of Legal Studies Vol. 10 1990, p. 460.
101 Hart, CL, supra note 7, at 198-9. See also Hart, EB, supra note 7, p. 265; Hart, CL II, supra

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 411

Of course, one can accept a rule for non-moral reasons. One can, for example,
accept a rule that one should spend time outdoors on sunny days for non-moral
reasons, since this is simply not a matter of morality. But can a judge really
accept the rule of recognition, the foundation of the legal system, for non-moral
reasons? I don’t think so. As I see it, a judge who thinks of himself as obligating
– as distinguished from simply coercing – a person to pay back a loan, or to
spend time in jail, etc., will have to accept the rule of recognition for moral
reasons, or, at the very least, pretend that he does.102 For as Joseph Raz has
convincingly argued, one can accept a personal rule for reasons of self-interest,
but one cannot coherently accept a rule that imposes obligations on others, such
as the rule of recognition, for other than moral reasons.103 One might, for
example, hold that sticking to the rule even when it yields bad results will yield
better consequences on the whole than a strategy of picking and choosing would.
This argument becomes even more persuasive when considering that on Hart’s
theory, duty-imposing rules differ from other rules, inter alia, by being more
important.104
Hart maintains, however, that inclusion of a moral component in the judges’
acceptance of the rule of recognition conveys an unrealistic picture of the way
judges conceive of their task of identifying and applying the law.105 According
to Hart, when a judge takes up his office he finds a “firmly settled practice of
adjudication,” which requires him to apply the legal norms identified by certain
criteria, a practice that determines the central duties of his office.106 In this
situation judges “are committed in advance in the sense that they have a settled
disposition to do this [that is, apply the rule of recognition] without considering
the merits of so doing in each case and indeed would regard it not open to them
to act on their view of the merits.”107
But Hart’s analysis does not show that the judge’s “settled disposition” is not
a moral disposition.108 My own view is that most judges have this disposition
because they consider the legal system on the whole to be worthy of moral
approval – if things would change for the worse they would gradually give up
that disposition.

note 96, p. 257.

102 R.A. Duff clearly believes that acceptance of the rule of recognition is a matter of moral
acceptance, since he maintains that to accept a rule for reasons of self-interest or habit is to
not accept it at all. For, he explains, ”such a relationship to the rules is defective by the
standards of the practice itself.” Duff, R. A., Legal Obligation and the Moral Nature of
Law, Juridical Review Vol. 25 1980, p. 72. The reason, he explains, is that the acceptance
of a rule involves accepting the moral values internal to the rule. Id. p. 70.
103 Raz, Joseph, Hart on Moral Rights and Legal Duties, Oxford Journal of Legal Studies Vol.
4:1 1983, p. 130-1; Raz, Purity, supra note 73, p. 454-5.
104 Hart, CL, supra note 7, p. 84-5. For an analysis of this issue, See Hill, Roscoe E., Legal
Validity and Legal Obligation, Yale Law Journal Vol. 80 1970, p. 61-7.
105 Hart, EB, supra note 7, p. 158.
106 Id. p. 158.
107 Id. p. 158-9.
108 See Postema, Gerald J., The Normativity of Law, in Issues in Contemporary Jurisprudence:
The Influence of H. L. A. Hart (Ruth Gavison, ed.) 1987, p. 94-5.

© Stockholm Institute for Scandianvian Law 1957-2010

412 Torben Spaak: Kelsen and Hart on the Normativity of Law

If, then, we assume that the rule of recognition is best understood as a moral
rule, what follows in regard to the normativity of law? My view is that the
normative force of the rule of recognition cannot be transmitted down to the
lower levels of the Stufenbau. For it does not follow from the fact, if it is a fact,
that the legal officials have a moral duty to apply all and only the rules identified
by the rule of recognition, that those rules impose (moral or non-moral)
obligations on the citizens. Although the legal officials must reasonably consider
the citizens to be obligated by a court-decision, say, it does not follow that the
citizens must view things this way, still less that the citizens are thus obligated.
This should be clear from a brief look at the existence conditions for a legal
system. On Hart’s theory, a legal system exists if, and only if, (i) the legal
officials adopt the internal point of view in regard to the rule of recognition, and
(ii) the citizens conform to the law.109 According to Hart, the voluntary
acceptance by the officials is necessary to create authority, without which the
law could not establish its coercive power.110 The citizens need not adopt the
internal point of view, however, and may have any of a number of reasons to
conform, including fear of punishment.111 Hence the law need not extend its
protections and benefits to all groups in a society.112 As Hart notes, we have to
pay a price for having a developed legal system, viz. that “the centrally
organized power may . . . be used for the oppression of numbers with whose
support it can dispense.”113 But what kind of authority could the law have over
the people it oppresses?
I conclude that whatever normative force the rules of the legal system may
have, it is not derived from the rule of recognition, and that therefore the very
idea of accounting for law’s normativity by means of an ultimate norm or rule is
misconceived. Any realistic account of the normativity of law has to be based on
the properties of the whole legal system.

5 The Trump Thesis and the Significance of Strictly Legal

Normativity

I have argued that both Kelsen and Hart attempted to account for the normativity
of law in the strictly legal sense, and that they succeeded in doing so. But I have
also argued that normativity in the strictly legal sense is problematic because it
seems to be impossible to combine it with the thesis that the law necessarily
claims to trump moral and other reasons for action. If this is so, we must make a
choice: (i) we might adopt the moral conception of law’s normativity instead, (ii)
we might reject the thesis that the law is intrinsically normative, or (iii) we might
reject or qualify the thesis that the law necessarily claims to trump moral and
other reasons for action? James Harris – who has analyzed Kelsen’s conception

109 Hart, CL, supra note 7, p. 59-60, 109-14.

110 Id. p. 196.
111 Id. p. 197.
112 Id. p. 196.
113 Id. p. 198.

© Stockholm Institute for Scandianvian Law 1957-2010

 Torben Spaak: Kelsen and Hart on the Normativity of Law 413

of normativity and who complains that it is pallid – recommends that we reject

the assumption that law is intrinsically normative:
Kelsen’s pallid normativity is contrived and artificial. His legacy is a
challenge to account for the normativity of the law in some other way…. The
solution suggested here is that we give up the search for any intrinsic connection
between legality and “ought,” whilst recognizing, as Hart and others have
shown, that normativity, in many important ways, hovers over the law.
Propositions of law assert the existence (or absence) of duties prescribed by the
law presently in force in some jurisdiction. They do not, in and of themselves,
assert that anything, from any point of view, ought to be done.114
I cannot accept Harris’ analysis, however. Harris rejects (what he refers to as)
the standard negative argument for the normativity of law, namely that
“statements to the effect that behaviour is legally obligatory cannot, without
change of meaning, be translated into statements about past, present or future
events.”115 But his argumentation is not convincing. He simply maintains that
propositions of law “do not, in and of themselves, assert that anything, from any
point of view, ought to be done.”116 As I see it, propositions of law – as
distinguished from propositions about law – do not assert anything, but simply
express norms.117 That is to say, they provide (but do not assert) that certain
things ought to be done. Harris’ characterization of propositions of law seems
rather to be true of propositions about law, but such propositions are irrelevant
to the problem about the normativity of law.
I recommend instead that we qualify the thesis that the law necessarily claims
to trump moral and other reasons for action in the following way. Instead of
saying that the law necessarily claims to trump moral and other reasons for
action in the sense that it claims a right to coerce that entails a duty to obey, we
might say that the law necessarily claims to trump moral and other reasons for
action in the sense that it claims a right to coerce simpliciter.118 On this
understanding of the trump thesis, we can easily combine it with the strictly
legal conception of law’s normativity.
There remains one question to be considered. Even if we agree that the
strictly legal conception of law’s normativity can be combined with a qualified
version of the trump thesis, we may wonder whether the normativity of law thus
conceived is an important characteristic of the law. Hart, as we have seen,
rejected Austin’s theory of law, inter alia, on the ground that it could not
account for the distinction between being obligated and being obliged. But why
was Hart so concerned with this distinction? If he wasn’t trying to elucidate the
difference between authority and power, what was he trying to do? Stanley
Paulson suggests that on this issue, Hart’s quarrel with Austin concerned the

114 Harris, Normativity, supra note 80, p. 115.

115 See id. p. 110.
116 See the quotation above.
117 I have in mind here Hart’s distinction between statements of the law and statements about
the law, and I assume that Harris has the same distinction in mind. See Hart, EB, supra note
7, p. 144-5.
118 For more on this topic, See Soper, Deference, supra note 15, Ch. 3.

© Stockholm Institute for Scandianvian Law 1957-2010

414 Torben Spaak: Kelsen and Hart on the Normativity of Law

nature of facts: whereas Austin spoke of habitual obedience, Hart spoke of social
rules. Says Paulson:
…Hart’s quarrel with the empirico-reductive tradition is over the nature of
facts. Austin’s facts – for example, that of habitual obedience – lend themselves
to explication apart from a rule-governed scheme, whereas Hart’s facts – the
social fact, for example, that certain ultimate criteria of legal validity are
accepted in this or that legal system – do not.119
If Paulson is right, Hart has offered a new and perhaps more sophisticated
analysis of the type of social situation in which we find obligations. But we may
still ask why it was so important to Hart to distinguish between these two types
of social situation. I have not been able to find a satisfactory answer to this
question, and I doubt that there is one to be found. I therefore conclude that the
strictly legal conception of law’s normativity is not so important a characteristic
of law as Kelsen and Hart seem to have thought.
If one accepts this conclusion, one is likely to start thinking about the relation
between Kelsen’s and Hart’s theories, on the one hand, and Austin’s and
Bentham’s theories, on the other. Did Kelsen and Hart advance our
understanding of law as much as is commonly thought? I am not convinced that
this is the case – at least in regard to the question of law’s normativity. But a
fuller investigation of this difficult issue will have to await another occasion.

119 Paulson, Stanley L., Continental Normativism and Its British Counterpart: How Different
Are They?, Ratio Juris Vol. 6:3 1993, p. 240-1.

© Stockholm Institute for Scandianvian Law 1957-2010